Reduction of Nitroaromatic Pesticides with Zero-Valent Iron

Chemosphere 54 (2004) 255–263 www.elsevier.com/locate/chemosphere

Reduction of nitroaromatic pesticides with zero-valent iron

Young-Soo Keum, Qing X. Li *

Department of Molecular Biosciences and Bioengineering, University of Hawaii, 1955 East-West Road, Ag Sci 218, Honolulu, HI 96822, USA Received 5 February 2003; received in revised form 4 June 2003; accepted 4 August 2003

Abstract

Reduction of eleven nitroaromatic pesticides was studied with zero-valent iron powder. Average half-lives ranged from 2.8 to 6.3 h and the parent compounds were completely reduced after 48–96 h. The di-nitro groups of the 2,6- dinitroaniline herbicides were rapidly reduced to the corresponding diamines, with a negligible amount of partially reduced monoamino or nitroso products. Low levels of de-alkylated products were observed after 10 days. The nitro group of the organophosphorus insecticides was reduced dominantly to the monoamines but in a slower rate than the 2,6-dinitroanilines. A trace amount of oxon products was found. Reduction of nitro to amino was also the predominant reaction for the diphenyl ether herbicides. Aromatic de-chlorination and de-alkylation were minor reactions. These amine products were more stable than the parent compounds and 60% or more of the amines were detected after two weeks. Humic acid decreased the reduction rates of pendimethalin, and dichlone (a known quinone redox mediator) counteracted the eﬀect of humic acid on the reactivity. Storage of iron powder under air decreased the reactivity very rapidly due to iron oxidation. Repeated use of iron powder also showed similar results. The reduced activity of air- oxidized iron was recovered by purging with hydrogen, but not nitrogen. Integration of iron powder with hydrogen- and quinone-producing microbial technologies may be a viable mean for remediation of highly oxidized xenobiotics in the environment. 2003 Elsevier Ltd. All rights reserved.

Keywords: Iron; Nitroaromatics; Pesticide; Quinone; Reduction; Remediation

1. Introduction phenols (Kim and Carraway, 2000), nitroaromatics (Hundal et al., 1997; Klausen et al., 2001), and poly- Reductive technologies have been extensively studied chloroalkane (Loraine, 2001). Various types of reducing for the removal of highly oxidized chemicals, particu- agents exist in the environment, especially in anaerobic larly polyhalogenated chemicals from contaminated sediment and soil (Kao et al., 2001). Studies were done sites. Those chemicals include polychlorinated dibenzo- with natural and synthetic reductants such as zero- p-dioxins (PCDDs), polychlorinated biphenyls (PCBs) valent metal (Adriaens et al., 1996; Arienzo et al., 2001), (Adriaens et al., 1996; Burris et al., 1996; Wang and metal sulﬁde (Butler and Hayes, 1998; Loch et al., 2000), Zhang, 1997; Woods et al., 1999; Kao et al., 2001), quinone (Adriaens et al., 1996; Kao et al., 2001) and pesticides (Sayles et al., 1997; Eykholt and Davenport, vitamin B12 (Burris et al., 1996; Glod et al., 1997). 1998; Ghauch et al., 1999, 2001; Ghauch and Suptil, Iron (Fe2þ or Fe3þ)-catalyzed Fenton reaction is a 2000; Loch et al., 2000; Ghauch, 2001), halogenated widely accepted remediation technology to decompose recalcitrant pollutants including PCDDs (Kao and Wu, 2000; Arienzo et al., 2001). Zero-valent iron is also * Corresponding author. Tel.: +1-808-956-2011; fax: +1-808- employed in many environmental reclamation projects as 956-3542. a permeable reactive barrier (US EPA, 1997). Zero- E-mail address: [email protected] (Q.X. Li). valent iron-based reductive methods can be applicable to

0045-6535/$ - see front matter 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2003.08.003 256 Y.-S. Keum, Q.X. Li / Chemosphere 54 (2004) 255–263 non-halogenated pesticides (Ghauch, 2001; Ghauch pendimethalin and trifluralin, (b) diphenyl ether her- et al., 2001), PCDDs (Kluyev et al., 2002) and azo dye bicides of bifenox, nitrofen and oxyfluorfen, and (c) (Nam and Trantnyek, 2001). However, corrosion of organophosphorus insecticides of fenitrothion and zero-valent iron produces various kinds of oxide, hydr- methyl parathion (Fig. 1). The purity of these pesticides oxide, ferrous and/or ferric ions. Under oxygen-limited was 99% or higher. 2,3-Dichloro-1,4-naphthoquinone conditions such as groundwater, sediment and/or deep (dichlone, 98% purity) was obtained from the US En- soil, hydrogen and electron are produced with ferrous vironmental Protection Agency (US EPA) (Triangle ion (Arienzo et al., 2001) and act as reductants. Though Park, NC). Powdery iron (99%, <325 mesh) and ferrous the resulting iron oxide or ion can enhance the contam- sulfate were purchased from Aldrich (Milwaukee, WI). inant degradation via Fenton reaction (He et al., 2002), Iron (III) hydroxide (a-FeO2H, goethite) was obtained formation of oxide layer generally prevents the adsorp- from Alfa Aesar (Ward Hill, MA). Acetonitrile and tion of contaminants on fresh iron surface and results in chloroform were from Fisher Scientific (Pittsburgh, an inhibition of reduction reaction. Acid washing is PA). generally used to remove such an oxide and hydroxide layer (Lavine et al., 2001), but the recovery of reductive 2.2. Reduction with zero-valent iron powder activity of oxidized iron surface is not well understood. The reactivity is often determined by surface-related Phosphate buffer (20 ml, potassium phosphate, pH properties including specific surface area of iron (Wang 6.5, 100 mM) was fortified with the pesticides dis- and Zhang, 1997) and adsorption coefficient of con- solved in acetonitrile at a final concentration of 10 taminants (Kim and Carraway, 2000) because the reac- lM, followed by addition of zero-valent iron powder tion occurs on the surface. Compounds that adsorb on (1 g). the iron surface competitively with contaminants can After hydrogen purging for 5 min, the reaction vial affect the reductive removal of contaminants. Humic was sealed tightly and stirred at 35 C. Control exper- acid, a ubiquitous natural organic material (NOM), can iments were done with the same concentration of pes- be adsorbed easily on iron surface, and consequently, ticides (10 lM) in phosphate buffer without iron inhibits contaminant remediation (Trantnyek et al., powder. In each sampling event, the contents in the vial 2001). Humic acid may also enhance the reaction via a were extracted with chloroform (3 · 10 ml). After re- redox mediation mechanism (Dunnivant et al., 1992). moval of the organic solvent, residues were re-dissolved The effects of humic acid on the reduction of nitro- in acetonitrile (1 ml) and were analyzed with high per- aromatic compounds were not well studied. Simple formance liquid chromatography (HPLC). An aliquot nitroaromatic compounds were reduced rapidly with of the sample was analyzed with gas chromatography– zero-valent iron to the anilines as a stable product mass spectrometry (GC–MS). The products were ten- (Klausen et al., 2001). Parathion, having a nitrophenyl tatively identified by comparing of resulting mass group, was also reduced rapidly with zero-valent iron spectra with NIST/EPA/NIH mass spectral library (Ghauch et al., 1999), however, reduction products were (NIST). not identified. It is important to understand the effects of environmental factors such as humic acid, atmospheric oxidation or repeated use of iron powder on the reduc- 2.3. Effects of humic acid and quinone on reduction tion processes and kinetic rates, and to identify reaction intermediates and products. Few studies have been done Humic acid (20 mg l1) was dissolved in phosphate on the reductive removal of nitroaromatic pesticides. In buffer (20 ml, pH 6.5, 100 mM), followed by addition of this study, 11 pesticides, for which reductive remediation zero-valent iron powder (1 g). 2,6-Dinitroaniline pesti- studies were not found, were selected to represent the cides were subsequently added at a concentration of 10 wide structural diversity and reactivity of nitroaromatic lM. After hydrogen purging for 5 min, samples were pesticides. Effects of humic acid, a quinone mediator, capped tightly and stirred at 35 C. In the case of ben- hydrogen purging, and atmospheric oxidation and re- fluralin, additional experiments were done in different peated use of iron powder on the zero-valent iron- orders of reagent additions. The pesticide standard so- catalyzed reduction were studied. lution was added first to phosphate buffer and sat for 5 min at 35 C, followed by humic acid (a final concentration of 10 lM and 20 mg l1 for benfluralin and 2. Materials and methods humic acid, respectively). In the case of quinone-mediated reactions, dichlone (a final concentration of 20 2.1. Reagents mg l1) was added to a mixture of humic acid (20 mg l1) and iron powder (1 g/20 ml) in phosphate buffer, fol- The 11 pesticides were (a) 2,6-dinitroaniline herbi- lowed by pendimethalin (10 lM) addition. The controls cides of benfluralin, ethalfluralin, nitralin, oryzalin, did not contain humic acid. Y.-S. Keum, Q.X. Li / Chemosphere 54 (2004) 255–263 257

R R 1 N 2 R2 S O NNO R1 O O 2 2 O R1 P CH3 O R O2N CH3 3 R3 NO2 R4

2,6-Dinitroaniline Diphenyl ether Organophosphorus

Class Name Substituent

R1 R2 R3 R4

2,6-Dinitroaniline Benfluralin (CH2)3CH3 CH2CH3 H CF3

Ethalfluralin CH2C(CH3)=CH2 CH2CH3 H CF3

Nitralin (CH2)2CH3 (CH2)2CH3 H SO2CH3

Oryzalin (CH2)2CH3 (CH2)2CH3 H SO2NH2

Pendimethalin CH2CH(C2H5)CH3 H CH3 CH3

Trifluralin (CH2)2CH3 (CH2)2CH3 H CF3

Diphenyl ether Bifenox CO2CH3 Cl Cl Nitrofen H Cl Cl

Oxyfluorfen OCH2CH3 Cl CF3

Organophosphorus Fenitrothion CH3 Me-parathion H

Fig. 1. Structures of eleven nitroaromatic pesticides.

2.4. Repeated use and storage of iron powder under bated for 96 h at 35 C. Hydrogen or nitrogen gas was atmospheric condition purged through the samples for 5 min and re-incubated. The control contained iron hydroxide suspen- Pendimethalin standard solution was added to sion (50 g l1) or ferrous sulfate (10 mM) instead of phosphate buffer (20 ml, pH 6.5, 100 mM) at a final iron powder. concentration of 10 lM, followed by zero-valent iron (1 g). The reaction mixture sat for 48 h, and then, the 2.6. Instrumental analysis iron powder was filtered and washed with degassed acetonitrile (2 · 100 ml) and phosphate buffer (100 ml). Reaction and products were monitored with HPLC Pendimethalin was added again after re-suspension of and GC–MS. The HPLC system consisted of a Shimadzu the iron powder in phosphate buffer. Pendimethalin LC-10S pump, Perkin-Elmer UV/VIS detector and a degradation was monitored after another 48-h incuba- VyDac C18 column (250 mm, 4.6 mm, 5 lm). The mobile tion. phase was 70% aqueous acetonitrile at a flow rate of 0.6 Fresh iron powder was exposed to air for 3, 7 and 10 or 0.8 ml min1 varying with analytes. The detection days at 35 C to examine effects of atmospheric oxida- wavelength was 210 nm. tion on the reduction reaction. An aliquot amount of the The GC–MS consisted of a Varian CP-3800 gas air-exposed iron (1 g) was added into pendimethalin chromatograph, Saturn 2000 ion trap MS, Varian CP- solution (20 ml, 10 lM in 100 mM-phosphate buffer, pH 8400 autosampler, and a ZB-1 column (60 m, 0.25 lm 6.5) and incubated at 35 C. film thickness, Phenomenex, USA). The carrier gas was helium at a flow rate of 2 ml min1. The MS was oper- 2.5. Hydrogen and nitrogen purging ated in electron impact mode (70 eV). Injection port and ion source temperatures were 250 C and the column Iron powder (50 g l1) that was exposed to air for temperature started at 35 C for 2 min, was jumped up 10 days was poured into pendimethalin solution (10 to 280 C at a rate of 4 C min1 and was finally held for lM in 100 mM-phosphate buffer, pH 6.5) and incu- 10 min. 258 Y.-S. Keum, Q.X. Li / Chemosphere 54 (2004) 255–263

Table 1

Reduction constants (k) and half-lives (t1=2) of nitroaromatic pesticides with zero-valent iron in presence or absence of humic acid 1;a a Chemical class Chemical name k,h t1=2,h Humic acid free Humic acidb Humic acid free Humic acidb 2,6-Dinitroaniline Benfluralin 0.19 ± 0.02 0.15 ± 0.03 3.6 ± 0.5 4.6 ± 1.4 0.20 ± 0.05c 3.5 ± 1.3b Ethalfluralin 0.18 ± 0.03 0.08 ± 0.02 3.9 ± 0.9 8.3 ± 3.3 Nitralin 0.16 ± 0.01 0.02 ± 0.00 4.3 ± 0.4 34.8 ± 0.0 Oryzalin 0.16 ± 0.03 0.04 ± 0.00 4.3 ± 1.2 17.3 ± 0.0 Pendimethalin 0.13 ± 0.01 0.01 ± 0.00 5.3 ± 0.6 69.3 ± 0.0 Trifluralin 0.13 ± 0.02 0.03 ± 0.00 5.3 ± 1.2 23.1 ± 0.0 Diphenyl ether Bifenox 0.16 ± 0.04 4.3 ± 1.6 Nitrofen 0.25 ± 0.01 2.8 ± 0.2 Oxyfluorfen 0.12 ± 0.02 6.0 ± 1.4

Organophosphate Fenitrothion 0.13 ± 0.01 5.3 ± 0.6 Methyl parathion 0.11 ± 0.02 6.3 ± 1.7

Quinoned Dichlone 0.01 ± 0.00 0.010 ± 0.00 69.3 ± 0.0 69.3 ± 0.0 a Average values ± standard deviation, values of three separate experiments. b The concentration of humic acid was 20 mg l1. c Benﬂuralin was added prior to the addition of humic acid. d Dichlone is used as redox mediator in the study of the eﬀect of humic acid.

3. Results and discussions The result showed that the reductive aromatic dechlorination was much slower than the reduction of 3.1. Reduction of nitro group by zero-valent iron powder nitro groups. Such a stability of aromatic chlorine to iron-mediated reduction was reported with p; p0-DDT All the pesticides in phosphate buffer were decom- (Sayles et al., 1997). The nitro group of fenitrothion and posed rapidly by zero-valent iron [kðrateÞ¼0:11 to methyl parathion was labile to reduction and the cor- 1 )0.25 h , half-lives ðt1=2Þ¼2:8–6:3 h] (Table 1). In all responding amines were the most abundant reaction the cases, the reaction followed pseudo-first order ki- products (>90%). Small quantities of the oxon and netics. Although an initial lag phase was reported (La- amino-oxon derivatives were found during the first 48 h vine et al., 2001), it was not observed in this study, (<1%). All the oxon products disappeared after 72 h. It which may be attributed to strong hydrophobicity of is known that the oxon products are more toxic than the the test chemicals and subsequent rapid adsorption on parent organophosphorus insecticides. Therefore, oc- iron surface. Reduction of the two organophosphates currence of the oxon products should be monitored in was generally slower than the 2,6-dinitroaniline herbi- iron-mediated reactions. The iron-free reaction (control) cides. It appeared that the methyl group adjacent to produced a similar amount of oxon as the iron mediated the nitro group in fenitrothion slightly enhanced the reaction, which suggested that the toxic oxon was not reduction in relation to its close analog methyl para- produced from zero-valent iron catalysis. However, the thion. However, the electron withdrawing groups of the oxon products were more resistant to degradation in methoxycarbony group in bifenox and the ethoxy in control samples and found even after 72 h. oxyfluorfen lowered the reaction rate in comparison The two nitro groups of the 2,6-dinitroaniline her- with nitrofen that has no substitution group adjacent to bicides were reduced readily to yield the corresponding the nitro group. 2,6-diamines as exemplified in Fig. 2. The monoamino The GC–MS results showed that reduction of nitro- products from the diphenyl ethers and organophos- to-amino was a dominant reaction for all the tested phates and di-amino derivatives from the 2,6-dinitro- compounds (Table 2). The nitro group of the diphenyl anilines were fairly stable relative to their corresponding ether herbicides was completely reduced to amino within parent compounds. These amino products had a t1=2 48 h and the corresponding amines persisted even after of longer than two weeks. Though the nitro groups 96 h. Trace quantities of dechlorinated amino products can be reduced stepwise [Eq. (1)] (Klausen et al., were found after 72 h and continuously increased during 2001), only trace quantities of the intermediates were the course of the experiment. found. Y.-S. Keum, Q.X. Li / Chemosphere 54 (2004) 255–263 259

NO2 NO NHOH NH2 2e-, 2H+ 2e-, 2H+ 2e-, 2H+ ð1Þ

H2OH2O

Monoamino and nitroso products were found in the (Table 1). In addition to inhibitory effects of the com- initial reaction mixtures of benfluralin, pendimethalin petitive binding, humic acid and related NOM can form and trifluralin. The monoalkyl nitrobenzimidazole was micelles and consequently trap the contaminants in the also found in the initial reaction stage of trifluralin (Fig. micelles (Trantnyek et al., 2001), which can limit 2), but it disappeared after 72 h. It may be formed by an the actual amount of contaminants to react with iron. intra-molecular ring closure from the partially reduced The rapid reaction of benfluralin (k ¼0:20 h1), when products. it was treated prior to the addition of humic acid, suggested that the adsorption of the contaminants is critical for the reduction. The inhibitory effect of humic acid can 3.2. Effects of humic acid and quinone be a serious problem for the application of zero-valent iron in the environment since NOM is ubiquitously All the above iron-mediated reactions were per- distributed in the environment. Therefore, methods to formed in phosphate buffer without any other additives. minimize the inhibitory effect of humic acid and NOM Environmental matrices contain various solutes includ- or to enhance the reaction should be studied. ing different inorganic salts and humic materials. Humic Opposite to humic acid, several natural quinones acid and related NOMs are ubiquitous in soil, sediment, (constituents of humic materials) are known to enhance and natural waters and exist on soil minerals or in reduction by a rapid quinone–hydroquinone intercon- aqueous solutions. Because of its strong adsorption on version (Weber, 1996; Trantnyek et al., 2001). Dichlone iron surface, NOM can affect the reduction with zero- (a known quinone redox mediator) enhanced the iron valent iron (Trantnyek et al., 2001). It can inhibit the mediated reduction rate of pendimethalin with humic reaction by competitive surface adsorption (Dunnivant acid from k ¼0:01 h1 to k, )0.12 h1, close to those et al., 1992; Curtis and Reinhard, 1994; Trantnyek et al., of the dichlone-iron treatment (k, )0.12 h1) and iron 2001). It may also enhance the reaction by electron control (k, )0.13 h1) (Fig. 3, and Table 1). The partial shuttle mechanisms through quinones (Dunnivant et al., recovery may be due to (a) a slow diffusion of the redox 1992). mediator dichlone from iron surface to the target mol- In this study, humic acid decreased the reaction rate ecules, (b) a slow reaction between quinone and pendi- of the 2,6-dinitroaniline herbicides for 1.2–10 fold methalin or (c) a competitive binding of dichlone and

Table 2 GC–MS identification of major reaction products Chemical class Reactants Molecular ion and fragments of major a reduction product (R–NO2 fi R–NH2) 2,6-Dinitroaniline Benfluralin (335)b 275 (Mþ, 47)c, 256 (18), 246 (100) Ethalfluralin (333) 273 (Mþ, 63), 230 (100), 202 (72) Nitralin (345) 285 (Mþ, 38), 270 (27), 242 (37) Oryzalin (346) 286 (Mþ, 45), 270 (19), 206 (100) Pendimethalin (281) 221 (Mþ, 27), 205 (44), 149 (100) Trifluralin (335) 275 (Mþ, 100), 256 (52), 231 (7) Diphenyl ether Bifenox (342) 311 (Mþ, 57), 296 (23), 286 (100) Nitrofen (284) 253 (Mþ, 62), 183 (17), 108 (100) Oxyfluorfen (362) 331 (Mþ, 72), 286 (28), 195 (100) Organophosphorus Fenitrothion (277) 247 (Mþ, 100), 215 (6), 138 (74) Methyl parathion (263) 233 (Mþ, 100), 201(3), 124 (51) a The major products are the corresponding 2,6-diamino dinitroanilines, monoamino diphenyl ethers and monoamino organophosphates, which the rest of the molecules remains unchanged. b Number in parenthesis was molecular weight of parent pesticides. c Mþ, molecular ion; number in parenthesis is the percent relative abundance. 260 Y.-S. Keum, Q.X. Li / Chemosphere 54 (2004) 255–263

pendimethalin on iron surface. The pendimethalin reduction rate in the dichlone-iron treatment would be faster than that of the iron control if a simple additive eﬀect of dichlone on the reaction rate exists. A slight decrease of the pendimethalin reduction rate in the dichlone-iron treatment suggests a possible competitive binding between pendimethalin and dichlone on iron surface, which may result in a slow reaction. In addition, degradation of dichlone by zero-valent iron can be another cause of such non-additive nature. If the redox mediation was achieved without other reaction, namely dechlorination or polymerization, the concentration of dichlone should be kept constant. However, dichlone concentration decreased slowly (Table 1), which may result in loss of redox mediation activity. It should be pointed out that the dichlone-mediated reactions gave the same products and the according diamine being the most abundant product as the other treatments.

3.3. Eﬀects of repeated use and storage under air

Disadvantages of using zero-valent iron for remediation include gradual loss of reactivity and permeability. These phenomena occur mainly by the formation of iron oxide and hydroxide layer (US EPA, 1998; Phillips et al., 2000). Plugging by iron-oxidized products is a serious problem in many permeable reactive barrier technologies. In aerobic or partially aerobic aqueous solution, such an iron hydroxide (e.g., goethite) can inhibit further corrosion by which formation of hydrogen or electron (i.e., reductant) is also inhibited. In addition, an Fig. 2. Representative GC–MS total ion chromatogram of a iron hydroxide layer inhibits adsorption of contami- reaction mixture of trifluralin (A), and GC–MS spectra of tri- nants on fresh iron surface. High concentration of dis- fluralin (B), diaminotrifluralin (C) and monoalkylbenzimida- solved oxygen can compete as an electron acceptor with zole (D). I, II and III in A denote trifluralin, diaminotrifluralin contaminants (US EPA, 1998; Lavine et al., 2001). and monoalkylbenzimidazole, respectively. In this study, repeated use and air exposure of fresh iron powder were studied to correlate the reaction rate with the status of iron surface. The removal rate of pendimehtalin gradually slowed down as iron powder was repeatedly used (Fig. 4A). In addition to the formation of an iron hydroxide layer, accumulation of reduction products such as the aromatic diamines may retard the reaction since aromatic amine is known as corrosion inhibitor (Tan and Blackwood, 2003). In this study, iron powder was completely washed with organic solvent for the subsequent use. Therefore, the reduction products would not be accumulated on the iron surface. As the repeated exposure to new batches of pesticide solutions, surface iron can be oxidized by trace oxygen. The pendimethalin reduction rates by the first (fresh iron), second and third repeated uses of iron powder were 0.13, 0.06, and 0.02 h1, respectively. A negative correlation between storage period and reduction rate Fig. 3. Effects of humic acid and dichlone on zero-valent iron (Fig. 4B) suggested that the surface oxidation slowed reduction of pendimethalin. Humic acid and dichlone were 20 down the reaction. In many remediation projects, loss mg l1 and 20 mg l1, respectively in phosphate buffer. of long-term performance of a permeable reactive iron Y.-S. Keum, Q.X. Li / Chemosphere 54 (2004) 255–263 261

1st Fresh 100 2nd 3 days 100 3rd 10 days 75 75

50 50 % Unreacted % Unreacted 25 25 (A) (B) 0 0 0102002550 Time (h) Time (h)

Fig. 4. Repeated uses of zero-valent iron (A) and effects of atmospheric oxidation (B) of zero-valent iron on reduction of pendimethalin in phosphate buffer solution. barrier was caused by such an oxidation (US EPA, 1998; be stimulated (Weathers et al., 1997), hydrogen can be a Farrell et al., 2000; Phillips et al., 2000). possible candidate to improve the performance of zero- valent iron permeable reactive barrier or combined 3.4. Reactivity recovery for air oxidized iron powder using technologies of zero-valent iron and bioremediation. hydrogen gas Hydrogen gas purging restored lost reactivity of air- oxidized iron powder for the reduction of pendimethalin, Although hydrogen and electron are produced from by increasing the reaction rate from 0.003 to 0.16 h1, but the corrosion of surface iron, such corrosion and con- not nitrogen (Fig. 5A). But the rate enhancement was not comitant formation of oxide or hydroxide layer inhibit found with freshly prepared iron powder. Hydrogen further reaction. Studies were done to improve the per- purging without iron powder did not change the re- formance and duration of zero-valent iron reactive per- duction rate of pendimethalin in buffer, ferrous sulfate meable barrier. Those include combined treatment of solution, and iron hydroxide suspension (Fig. 5B). Nu- aluminosilicate mineral (Powell and Puls, 1997), qui- merous studies were done for the effects of solution pH nones (Curtis and Reinhard, 1994), or organic solvent on the zero-valent iron mediated reaction (Powell and and surfactant (Loraine, 2001). Integration of zero- Puls, 1997; Lavine et al., 2001), however, none were valent iron into different technologies was also studied, found on recovering the reactivity of oxidized iron by for example, iron and anaerobic bioremediation (Weath- hydrogen. Apparent recoveries of the reactivity by hy- ers et al., 1997; Gerlach et al., 2000) or oxidative reaction drogen may be explained by possible (a) reduction of (Krishna et al., 2001). Acid washing is often used to re- iron oxide and hydroxide to more reactive species or (b) move iron oxide and hydroxide from the surface (Lavine catalytic hydrogenation by purging hydrogen gas on the et al., 2001). Hydrogen can increase the dissolution rate active surface (Odziemakowski et al., 2000). Although of metals by which hydrogen or electron can be produced hydrogen gas was supplied artificially in this study, the more rapidly (Wallinder et al., 2001) and the growth of results suggest a possibility of using hydrogen producing anaerobic microorganism, having reductive activity, can microorganisms and zero-valent iron for removal of

Fig. 5. Effects of hydrogen and nitrogen purging on pendimethalin reduction by air-oxidized iron powder (A) and effects of hydrogen purging on pendimethalin reduction by iron hydroxide, ferrous sulfate and phosphate buffer as control (B). The start time of hydrogen and nitrogen purging was marked with an arrow. 262 Y.-S. Keum, Q.X. Li / Chemosphere 54 (2004) 255–263 highly oxidized pollutants. Hydrogen purging may be containing natural organic matter. Environ. Sci. Technol. applicable to remediation of nitroaromatic pesticides. 26, 2133–2141. Eykholt, G.R., Davenport, D.T., 1998. Dechlorination of chloroacetanilide herbicides alachlor and metolachlor by 4. Conclusion iron metal. Environ. Sci. Technol. 32, 1482–1487. Farrell, K., Kason, M., Melitas, N., Li, T., 2000. Investigation The nitroaromatic pesticides were rapidly reduced of long-term performance of zero-valent iron for reductive with zero-valent iron to the corresponding amines as dechlorination of trichloroethylene. Environ. Sci. Technol. major reduction products. Nitroso intermediates were 34, 514–521. found only at very small quantities in some reactions. Gerlach, R., Cunningham, A.B., Caccavo, F.J., 2000. Dissim- ilatory iron-reducing bacteria can influence the reduction of Dichlone, a quinone, neutralized the inhibitory effect of carbon tetrachloride by iron metal. Environ. Sci. Technol. humic acid on the catalysis ability of zero-valent iron for 34, 2461–2464. the reduction reaction of pendimethalin, which suggests Ghauch, A., 2001. Degradation of benomyl, picloram, and that some natural quinones be relevant to pollutant dicamba in a conical apparatus by zero-valent iron powder. degradation via iron-catalyzed reduction in the envi- Chemosphere 43, 1109–1117. ronment. Although use of zero-valent iron for environ- Ghauch, A., Suptil, J., 2000. Remediation of s-triazines mental remediation is economical and feasible, it is contaminated water in a laboratory scale apparatus using important to develop ways to enhance and stabilize its zero-valent iron powder. Chemosphere 41, 1835–1843. catalysis capability and recover the activity of oxidized Ghauch, A., Rima, J., Amine, C., Martin-Bouyer, M., 1999. iron. Hydrogen gas purging rapidly recovered the ca- Rapid treatment of water contaminated with atrazine and parathion with zero-valent iron. Chemosphere 39, 1309– talysis ability of air-exposed iron to reduce pendimeth- 1315. alin, which implies hydrogen purging as a possible mean Ghauch, A., Gallet, C., Charef, A., Rima, J., Martin-Bouyer, to maintain a reactive zero-valent iron barrier for the M., 2001. Reductive degradation of carbaryl in water by associated remediation technologies. It is possible to use zero-valent iron. Chemosphere 42, 419–424. zero-valent iron in conjunction with microbial ap- Glod, G., Angst, W., Holliger, C., Schwarzenbach, R.P., 1997. proaches that may provide hydrogen and quinones for Corrinoid-mediated reduction of tetrachloroethene, trichlo- reduction of highly oxidized pollutants. roethene, and trichlorofluoroethene in homogeneous aqueous solution: reaction kinetics and reaction mechanisms. Environ. Sci. Technol. 31, 253–260. Acknowledgements He, J., Ma, W.-H., He, J.-J., Zhao, J.-C., Yu, J.C., 2002.

Photooxidation of azo dye in aqueous dispersions of H2O2/ The work was supported in part by US EPA grant a-FeOOH. Appl. Catal. B 39, 211–220. no. 989512-01-1 and USDA/CSREES/TSTAR grant no. Hundal, L.S., Singh, J., Bier, E.L., Shea, P.J., Comfort, S.D., 0034135-9576. Powers, W.L., 1997. Removal of TNT and RDX from water and soil using iron metal. Environ. Pollut. 97, 55–64. Kao, C.M., Wu, M.J., 2000. Enhanced TCDD degradation by References Fenton’s reagent peroxidation. J. Haz. Mat. B 74, 197–211. Kao, C.M., Chen, S.C., Liu, J.K., Wu, M.J., 2001. Evaluation Adriaens, P., Chang, P.R., Barkovskii, A.L., 1996. Dechlori- of TCDD biodegradability under different redox conditions. nation of PCDD/F by organic and inorganic electron Chemosphere 44, 1447–1454. transfer molecules in reduced environments. Chemosphere Kim, Y.H., Carraway, E.R., 2000. Dechlorination of penta- 32, 433–441. chlorophenol by zero valent iron and modified zero valent Arienzo, M., Chiarenzelli, J., Scrudato, R., Pagano, J., iron. Environ. Sci. Technol. 34, 2014–2017. Falanga, L., 2001. Iron-mediated reactions of polychlori- Klausen, J., Ranke, J., Schwarzenbach, R.P., 2001. Influence of nated biphenyls in electrochemical peroxidation process solution composition and column aging on the reduction of (ECP). Chemosphere 44, 1339–1346. nitroaromatic compounds by zero-valent iron. Chemo- Burris, D.R., Delcomyn, C.A., Smith, M.H., Roberts, A.L., sphere 44, 511–517. 1996. Reductive dechlorination of tetrachloroethylene cata- Kluyev, N., Cheleptchikov, A., Brodsky, E., Soyfer, V.,

lyzed by vitamin B12 in homogeneous and heterogeneous Zhilnikov, V., 2002. Reductive dechlorination of polychlo- systems. Environ. Sci. Technol. 30, 3047–3052. rinated dibenzo-p-dioxins by zerovalent iron in subcritical Butler, E.C., Hayes, K.F., 1998. Eﬀects of solution composition water. Chemosphere 46, 1293–1296. and pH on the reductive dechlorination of hexachloroethane Krishna, M.V.B., Chandrasekaran, K., Karunasagar, D., by iron sulﬁde. Environ. Sci. Technol. 32, 1276–1284. Arunachalam, J., 2001. A combined treatment approach Curtis, G.P., Reinhard, M., 1994. Reductive dehalogenation of using Fenton’s reagent and zero valent iron for the removal hexachloroethane, carbon tetrachloride, and bromoform by of arsenic from drinking water. J. Haz. Mat. B 84, 229–240. anthrahydroquinone disulfonate and humic acid. Environ. Lavine, B.K., Auslander, G., Ritter, J., 2001. Polarographic Sci. Technol. 28, 2393–2401. studies of zero valent iron as a reductant for remediation of Dunnivant, F.M., Schwarzenbach, R.P., Macalady, D.L., 1992. nitroaromatics in the environment. Microchem. J. 70, 69– Reduction of substituted nitrobenzenes in aqueous solutions 83. Y.-S. Keum, Q.X. Li / Chemosphere 54 (2004) 255–263 263

Loch, A.R., Lippa, K.A., Carlson, D.L., Chin, Y.P., Traina, Trantnyek, P.G., Scherer, M.M., Deng, B., Hu, S., 2001. Effects S.J., Roberts, A.L., 2000. Nucleophilic aliphatic substitu- of natural organic matter, anthropogenic surfactants, and tion reactions of propachlor, alachlor, and metolachlor with model quinines on the reduction of contaminants by zero- 2 bisulfide (HS ) and polysulfides (Sn ). Environ. Sci. Tech- valent iron. Wat. Res. 35, 4435–4443. nol. 36, 4065–4073. US EPA, 1997. Permeable reactive subsurface barriers for the Loraine, G.A., 2001. Effects of alcohols, anionic and nonionic interception and remediation of chlorinated hydrocarbon surfactants oh the reduction of PCE and TCE by zero- and chromium(VI) plumes in ground water. EPA/600/F-97/ valent iron. Wat. Res. 35, 1453–1460. 008. Nam, S.K., Trantnyek, P.G., 2001. Reduction of azo dyes with US EPA, 1998. Permeable reactive barrier technologies for zero-valent iron. Wat. Res. 34, 1837–1845. contaminant remediation. EPA/600/R-98/125. Odziemakowski, M., Gui, L., Gillham, R.W., 2000. Reduction Wallinder, D., Hultquist, G., Tveten, B., Hornlund, E., 2001. of N-nitrosodimethylamine with granular iron and nickel- Hydrogen in chromium: influence on corrosion potential enhanced iron. 2. Mechanistic studies. Environ. Sci. Tech- and anodic dissolution in neutral NaCl solution. Corros. nol. 34, 3495–3500. Sci. 43, 1267–1281. Phillips, D.H., Gu, B., Watson, D.B., Roh, Y., Liang, L., Lee, Wang, C., Zhang, W., 1997. Synthesizing nanoscale iron S.Y., 2000. Performance evaluation of a zerovalent iron particles for rapid and complete dechlorination of TCE reactive barrier: mineralogical characteristics. Environ. Sci. and PCBs. Environ. Sci. Technol. 31, 2154–2156. Technol. 34, 4169–4176. Weathers, L.J., Parkin, G.F., Alvarez, P.J., 1997. Utilization of Powell, R.M., Puls, R.W., 1997. Proton generation by disso- cathodic hydrogen as electron donor for chloroform lution of intrinsic or augmented aluminosilicate minerals for cometabolism by a mixed, methanogenic culture. Environ. in situ contaminant remediation by zero-valence-state iron. Sci. Technol. 31, 880–885. Environ. Sci. Technol. 31, 2244–2251. Weber, E.J., 1996. Iron-mediated reductive transformations: Sayles, G.D., You, G., Wang, M., Kupferle, M.J., 1997. DDT, investigation of reaction mechanism. Environ. Sci. Technol. DDD, and DDE dechlorination by zero-valent iron. Envi- 30, 716–719. ron. Sci. Technol. 31, 3448–3454. Woods, S.L., Trobaugh, D.J., Carter, K.J., 1999. Polychlori-

Tan, C.K., Blackwood, D.J., 2003. Corrosion protection by nated biphenyl reductive dechlorination by vitamin B12s: multilayered conducting polymer coating. Corros. Sci. 45, thermodynamics and regiospeciﬁcity. Environ. Sci. Technol. 545–557. 33, 857–863.